The bleomycins (BLM's) are a family of antitumor antibiotics presently used clinically in the treatment of testicular cancer, head and neck carcinoma as well as Hodgkin's disease. Establishing the major intracellular target(s) and the mechanism(s) responsible for BLM's observed cytotoxicity and therapeutic efficacy is thus of great interest. BLM has been known for some time to effect both single strand (ss) and double strand (ds) breaks of DNA in vitro and in vivo. These ds breaks have been proposed to be the major contributing factor in BLM's cytotoxicity. A model for the way in which a single molecule of BLM can effect cleavage on two strands of DNA without dissociating has been proposed based on recent structural data acquired at CM[R. In this model, the bithiazole unit of the BLM molecule is thought to undergo a trans to cis flip that repositions the metal binding region of the molecule at the second strand. The hydroperoxide form of iron BLM (activated BLM) has been shown in mechanistic studies to abstract the 4'-H of the DNA ribose moiety. Once the 4'-H has been removed, two major DNA lesions are formed. A phosphoglycolate lesion can be formed in an oxygen dependent manner, and a 4'-keto abasic site can be formed by an oxygen independent pathway. The structure of the phosphoglycolate lesion may present the key to understanding the ds cleavage of the DNA by BLM. Previously, a phosphoglycolate lesion containing the GTAC sequence had been synthesized in our lab. The GTAC sequence was chosen since it is a hot spot for ds cleavage with a ratio of ds : ss cleavage of I : 3. The key to acquiring good data for this piece had been to synthesize the oligonucleotide as a double hairpin connected with hexaethyleneglycol spacers. The proton chemical shift assignments have been completed and modeling of the 2D NMR data collected on a 750 MHz NMR is currently underway. The oligonucleotide had also been titrated with a cobalt hydroperoxy form of BLM, which is a proposed analog of activated iron BLM. This titration will be repeated with a more concentrated sample in the near future. Preparation of this sample is underway. The complex of cobalt BLM and DNA may hold the key to understanding the mechanism of ds cleavage. It may be possible to cont rast the data from the phosphoglycolate lesion with the data from an intact piece of DNA and detect a trans to cis isomerization in the bithiazole region. Another oligonucleotide sample containing the phosphoglycolate lesion has been prepared as well. This oligonucleotide is also a double hairpin linked by hexaethyleneglycol spacers. However, this piece contains the GGCC sequence. Interestingly, this sequence cannot undergo ds cleavage via the BLM molecule. This sequence thus functions as a control experiment. Data for this DNA has been collected in D20 and 90%H20/10%D20. The assignments of the chemical shifts are in progress. This oligonucleotide will also be titrated with a cobalt hydroperoxy BLM. If this oligonucleotide yields a one to one complex with the BLM, it will be interesting to contrast this data with the data from the GTAC piece. The synthesis of the 4'-keto abasic site is also in progress in the lab. This lesion is a synthetic challenge and efforts to make this lesion have thus far been unsuccessful. However, a new approach to this problem has been taken. A 4'-azido-2'-deoxyuridine moiety has been synthesized that will be incorporated into DNA using a polymerase and a kinase. Subsequent reduction will yield the 4'-keto abasic site. Since the 4'-keto abasic site is currently unavailable for structure determination, a close relative, a 4'-OH abasic site is being studied by 2D NMR. Again, this data was acquired on the 750 MHz instrument at CMR. This abasic site is being studied in the GTAC sequence context for comparison with the phosphoglycolate lesion. The modeling of this duplex 13-mer containing the abasic site is in the final stages of refinement. Interestingly, two distinct conformations of this abasic site in the GTAC region are present in equal amounts. This leads to the question of recognition of this damage site by DNA repair enzymes. It is reasonable to postulate that one of these conformations is recognized preferentially by DNA repair enzymes such as human apurinic/apyrimidinic endonuclease (APEI). In summary, modeling of 2D NMR data acquired on the 750 MHz instrument are at various stages of refinement for the abasic site and the phosphoglycolate lesions. Assignments of the 2D NMR data are in progress for additional phosphoglycolate lesions with and without Co-BLM bound.
Showing the most recent 10 out of 12 publications
